Towards the successful deployment of 5G in Europe · 2020. 12. 14. · Centre on Regulation in Europe (CERRE) asbl Avenue Louise, 475 (Box 10) - B-1050 Brussels - Belgium Ph: +32

Centre on Regulation in Europe (CERRE) asbl

Avenue Louise, 475 (Box 10) - B-1050 Brussels - Belgium

Ph: +32 2 230 83 60 - Fax: +32 2 230 83 70 – [email protected] – www.cerre.eu

Towards the successful deployment of 5G

in Europe:

What are the necessary policy and regulatory

conditions?

Project report

Dr. Ir. Wolter Lemstra (CERRE, Delft University of Technology &

Nyenrode Business Universiteit)

Prof. Martin Cave (CERRE & Imperial College London)

Prof. Marc Bourreau (CERRE & Telecom ParisTech)

30 March 2017

170330_CERRE_5GReport_Final 2/116

Table of Contents

About CERRE ...................................................................................................................... 5

About the authors .............................................................................................................. 6

Executive Summary ............................................................................................................ 7

Acknowledgements .......................................................................................................... 10

1 Introduction ............................................................................................................. 11

2 The leadership role ................................................................................................... 14

2.1 Regularities in next generation mobile communication: 1G through 4G .................. 18

2.2 Interpretation of the regularities and trends leading to 5G ...................................... 22

3 5G architecture and features .................................................................................... 25

3.1 5G requirements ........................................................................................................ 25

3.2 5G architecture and virtualisation ............................................................................. 28

3.3 5G roadmap ............................................................................................................... 31

3.4 The demand-side perspective.................................................................................... 32

3.4.1 The market for connections and devices .............................................................. 32

3.4.2 The IoT market ...................................................................................................... 35

4 The two stylised images of the 5G future................................................................... 38

4.1 ‘Evolution’ and ‘Revolution’ ....................................................................................... 38

4.2 The industry structure ............................................................................................... 39

5 The ‘Evolution’ image ............................................................................................... 40

5.1 Anticipated outcome – attractiveness of the outcome ............................................. 40

5.2 Industry structure in the ‘Evolution’ image ............................................................... 42

5.2.1 Societal .................................................................................................................. 43

5.2.2 Economic ............................................................................................................... 43

5.2.3 Political and Regulatory ......................................................................................... 43

5.2.4 Technological ......................................................................................................... 44

5.2.5 Environmental ....................................................................................................... 45

5.2.6 Rivalry .................................................................................................................... 45


5.2.7 Barriers to entry .................................................................................................... 48

5.2.8 Substitutes ............................................................................................................. 48

5.2.9 Buyers and buyer power ....................................................................................... 49

5.2.10 Suppliers and supplier power ................................................................................ 49

5.2.11 Market structure in the ‘Evolution’ image ............................................................ 50

6 Policy and regulatory actions enabling the ‘Evolution’ image ..................................... 52

6.1 Policy action – the 5G Action Plan ............................................................................. 53

6.2 Regulatory actions ..................................................................................................... 55

7 The ‘Revolution’ image ............................................................................................. 59

7.1 Anticipated outcome – attractiveness of the outcome ............................................. 59

7.2 Outline of the ‘Revolution’ image .............................................................................. 62

7.3 Industry structure in the ‘Revolution’ image ............................................................. 68

7.3.1 Societal .................................................................................................................. 68

7.3.2 Economic ............................................................................................................... 68

7.3.3 Political and Regulatory ......................................................................................... 69

7.3.4 Technological ......................................................................................................... 69

7.3.5 Environmental ....................................................................................................... 70

7.3.6 Rivalry .................................................................................................................... 70

7.3.7 Barriers to entry .................................................................................................... 72

7.3.8 Substitutes ............................................................................................................. 73

7.3.9 Buyers and buyer power ....................................................................................... 73

7.3.10 Suppliers and supplier power ................................................................................ 74

7.3.11 Market structure in the ‘Revolution’ image .......................................................... 75

8 Policy and regulatory actions enabling the ‘Revolution’ image ................................... 77

8.1 Policy formation and implementation ....................................................................... 77

8.2 Policy actions ............................................................................................................. 78

8.2.1 5G Action Plans ...................................................................................................... 79

8.3 Regulatory actions ..................................................................................................... 80

9 Summary .................................................................................................................. 87


Annex A: Characteristics of the mobile communications business ...................................... 91

Annex B: Abbreviations and acronyms .............................................................................. 94

Annex C: Timeline of major mobile communication events ................................................ 99

Annex D: 5G and its spectrum requirements - an overview ............................................... 105

Annex E: 5G and net neutrality ......................................................................................... 110

References ...................................................................................................................... 115


About CERRE

Providing top quality studies and dissemination activities, the Centre on Regulation in Europe

(CERRE) promotes robust and consistent regulation in Europe’s network and digital industries.

CERRE’s members are regulatory authorities and operators in those industries as well as

universities.

CERRE’s added value is based on:

• its original, multidisciplinary and cross-sector approach; • the widely acknowledged academic credentials and policy experience of its team and

associated staff members;

• its scientific independence and impartiality; • the direct relevance and timeliness of its contributions to the policy and regulatory

development process applicable to network industries and the markets for their

services.

CERRE's activities include contributions to the development of norms, standards and policy

recommendations related to the regulation of service providers, to the specification of market

rules and to improvements in the management of infrastructure in a changing political,

economic, technological and social environment. CERRE’s work also aims at clarifying the

respective roles of market operators, governments and regulatory authorities, as well as at

strengthening the expertise of the latter, since in many Member States, regulators are part of a

relatively recent profession.

The project, within the framework of which this report has been prepared, has received the

financial support of a number of CERRE members. As provided for in the association's by-laws, it

has, however, been prepared in complete academic independence. The views expressed in this

CERRE report are those of the author(s). They do not necessarily correspond to those of CERRE,

to any sponsor or to any (other) member of CERRE.


About the authors

Wolter Lemstra is a CERRE Research Fellow, Senior Research Fellow at the Faculty Technology,

Policy & Management of the TUDelft, Associate Professor, Nyenrode Business Universiteit and

Senior Lecturer at the Strategy Academy, the Netherlands. His research interests are the

developments of the telecommunication sector in relation to firm strategy and government

policy, and the role of governance regimes and the institutional environment. He thereby links

his academic interests to 25 years of experience in the telecom sector. He occupied senior

management positions in the field of engineering and product management, sales and

marketing, strategy and business development. Most recently he was a Member of the Senior

Management Team and Vice-President at Lucent Technologies, responsible for marketing and

business development in the Europe, Middle East and Africa region.

Martin Cave is Joint Academic Director at CERRE. He is a regulatory economist specialising in the

regulation of network industries, especially the communications sector. He is currently a visiting

professor at Imperial College Business School, having formerly held chairs at Brunel University

(in the Department of Economics), at Warwick University (in the Business School), and at LSE (in

the Law Department). He has written a number of books and papers on aspects of

communications regulation, including Spectrum Management: Using the Airwaves for Maximum

Social and Economic Benefit (Cambridge University Press, 2015), co-authored with William

Webb.

Marc Bourreau is a Joint Academic Director of CERRE, Professor of Economics at Telecom

ParisTech, and director of the Innovation & Regulation Chair at Telecom ParisTech. He is also

affiliated with the interdisciplinary institute for innovation (i3) for his research. Marc graduated

in engineering from Telecom ParisTech in 1992. He received his doctorate in economics from

University of Paris 2 Panthéon-Assas in 1999, and a “Habilitation à Diriger des Recherches” from

University of Paris 1 Panthéon-Sorbonne in 2003. From 1997 to 2000, he worked as a regulatory

economist at France Telecom/Orange. He became assistant professor at Telecom ParisTech in

2000. Marc has published widely in leading economics journals. He is Co Editor-in-Chief of

Information Economics & Policy, and a member of the editorial boards of the Review of Network

Economics, Telecommunications Policy and the DigiWorld Economic Journal (formerly

Communications & Strategies). He is also a member of the scientific committee of the Florence

School of Regulation at the European University Institute in Florence (Italy), an associate

researcher of the Laboratory of Industrial Economics (LEI), and an associate researcher of

Cepremap. His main research interests are in industrial organisation, regulation,

telecommunications, and digital economics.


Executive Summary

Historical regularity suggests that approximately every 10 years a new generation of mobile

communications technology is introduced. The next generation – 5G – is expected to be

introduced around 2020. Each new generation represents a complex interplay between

interdependent stakeholders, including infrastructure equipment manufacturers, device makers,

operators, and end-users, as well as regulators and policy makers at national, regional and

global level. This is a high-stakes game requiring deep investments which can only be successful

if well coordinated, and when supply and demand can be aligned.

European policy makers have a keen interest in the success of the next generation because

ubiquitous and high capacity electronic communication infrastructure is recognised as a

cornerstone of economic development and productivity growth. The second generation, GSM,

was a big success. It reached its peak in deployment in 2015 with 3.83 billion subscribers served

through over 700 operators in 219 countries and territories.

With 5G rapidly shaping up in the R&D and standardisation environment, what are the lessons

to be learned from 1G through 4G that should be taken into account to ensure a successful

development and deployment of 5G in Europe? What does 5G have in common with previous

generations and where is it different? What are the implications? Moreover, is the path towards

the future predetermined by the previous generations, by a prevailing industry structure, or are

there alternative routes? Is there possibly a fork in the road ahead that requires special

attention from policy makers and regulators, as it may lead to different futures? When there are

different futures with different outcomes, is one more desirable than the other? In sum, what

would be the policy and regulatory framework required to enable the success of 5G in Europe?

To respond to these questions, this report identifies first, on the basis of an assessment of the

previous generations of mobile communication technologies and against the backdrop of

European leadership in the development and deployment of GSM, the policy and regulatory

lessons to be drawn from the latter’s success.

Secondly, it provides a description of 5G, the performance objectives that have been assigned

to it, the latter’s architecture and key features; the report then compares those features with

previous generations.

Thirdly, it describes two stylised, extreme images of possible futures of 5G, ‘Evolution’ and

‘Revolution’. Those images represent two different sets of outcomes that are enabled by two

different sets of policies and regulatory interventions. They constitute a fork in the road that

policy makers and regulators will have to navigate in the years to come.

It should be emphasised that the latter do not aim to represent the complexity of how the

actual future may unfold, nor should they be considered as scenarios, such as those initiated by


Shell in the eighties. They are merely intended to stimulate the debate on the policy and

regulatory conditions for the successful deployment of 5G in Europe.

Fourthly, the report describes the policy and regulatory framework that would be required to

enable each of these images.

‘Evolution’ follows the pattern of previous generations and current trends. ‘Revolution’

represents a clear break with these trends. It exploits the opportunities of standardised

application programming interfaces (APIs) for service creation, being enabled by network

virtualisation as an architectural foundation of 5G. These open APIs allow the market entry of a

multitude of virtual mobile network operators (VMNOs). VMNOs are dedicated to serve

particular industry verticals or economic sectors with tailored feature sets and tailored qualities

of services.

In ‘Evolution’, the regularities and trends that can be observed from the previous generations of

mobile communication, i.e. 1G through 4G, are considered as the main determinants of the 5G

future. A key assumption in this image is that the core business of the mobile operators

continues to be serving the mass market of consumers.

‘Revolution’ reflects the shift to a layered model with multiple specialised providers at each

layer. At the lower layer are the passive infrastructure facilities providers. At the next layer up

are the network operators – the owners of radio frequency licenses and of active infrastructure

facilities. These mobile network operators are the wholesale providers of a range of connectivity

services with various grades of quality to the virtual mobile network operators (VMNOs) at the

top layer.

These VMNOs can be compared to the MVNOs of earlier generations, serving specific market

segments and leveraging a particular brand. However, they are different as VMNOs have full

control of a virtual slice of the network infrastructure to deliver services with differentiated

quality levels. In ‘Revolution’, the number of VMNOs is very large. In principle, each firm that

wishes to extend its reach to end-users through a mobile service can do so as a VMNO using its

own brand and applying bundling with other business services. As firms compete for end-users,

they are expected to compete for providing the best virtual mobile services as well. This results

in a very dynamic wholesale market. This is a market that unlocks a higher willingness to pay,

which, through differentiation of network services, will flow through to incentivise 5G network

investments.

The policy and regulatory actions that enable ‘Evolution’ build on the new Electronic

Communications Code and the 5G Action Plan. They are also related to the topics of, amongst

others, trading in radio spectrum usage rights, coverage obligations, indoor access, network

sharing, net neutrality and minimum requirements for public protection and disaster relief.

The policy and regulatory actions that enable ‘Revolution’ also build on the new Electronic

Communications Code and the 5G Action Plan. However, they also involve a 5G Action Plan

focused on the European-wide use of open APIs. The transition to the new industry


configuration is recognised as a major innovation project requiring restraint in terms of

regulation. Regulatory action is based on intervention only in case of market failure, e.g. in areas

such as retail market access, open and common APIs and national roaming. Special action is

required for net neutrality, liberalisation of SIM usage and use of multiple VMNOs on a single

device.


Acknowledgements

The authors would like to acknowledge the very valuable contributions received during the

research process from Peter Anker, Senior Research Fellow at Technology, Policy &

Management Department of the TU Delft and expert in the field of radio spectrum governance;

Herbert Ungerer, co-author of the EC Green Paper that started the liberalisation process of the

telecoms services sector in Europe, and who was also involved in the GSM frequency directive;

William Webb, Centre for Science and Policy, University of Cambridge; participants in the

workshop on 5G held at the department of Electrical Engineering of the TU Eindhoven; Nur

Engin at NXP; Jordi Domingo, Universitat Politècnica de Catalunya; Jorgen Abild Andersen, OECD;

participants in the CRplatform.NL workshop on 5G and on the Use Case of Academic Medical

Centers; the participants in the Round Table discussions organised by the Dutch Ministry of

Economic Affairs to shape the strategic agenda for mobile communications; Wessel Blom at

Verizon; Bert Dorgelo, former standards expert at Philips and participant in the GSM project;

and Wouter Franx and Anne van Otterlo at Nokia (formerly Alcatel-Lucent).


1 Introduction

Historical regularity suggests that approximately every 10 years a new generation of mobile

communications technology is introduced. The sequence started with 1G in 1981 and the latest

generation, 4G, was introduced in 2009. Hence, the next generation – 5G – is expected to be

introduced around 2020. Each new generation represents a complex interplay between

interdependent stakeholders, including infrastructure equipment manufacturers, device makers,

operators, and end-users, as well as regulators and policy makers at national, regional and

global level. The interplay concerns the allocation and assignment of new radio frequency

bands, the development of a new standard, the development of new network equipment, the

investment in new infrastructure build-out, the launch of new devices and the uptake by end-

users. This is a high-stakes game requiring deep investments which can only be successful if

well-coordinated, and when supply and demand can be aligned.

European policy makers have a keen interest in the success of the next generation because

ubiquitous and high-capacity electronic communication infrastructure is recognised as a

cornerstone of economic development and productivity growth. Moreover, at the European

level, electronic communications has become a strategic element in the creation of the single

internal market. Following the success of the second generation – GSM – the question of

European leadership in the development and deployment of cellular communications is being

raised with each successive generation.

The benchmark for European leadership in mobile communications is GSM, a second generation

technology introduced in 1991, which reached its peak in deployment in 2015 with 3.83 billion

subscribers served through over 700 operators in 219 countries and territories. This is

phenomenal achievement, especially when recognising that the nearest competing 2G

technology – CDMA – reached its peak with 374 million subscribers also in 2015. This represents

a factor 10 difference. However, in Europe the next generation 3G – UMTS is generally

considered as less successful, having had a slow start in deployment compared to a much faster

uptake of 3G in the USA and Asia. Nonetheless, from a consumer welfare perspective, 3G and

4G can be considered as quite successful, considering the price levels and the data rates

provided.

Therefore, with 5G rapidly shaping up in the R&D and standardisation environment, what are

the lessons to be learned from 1G through 4G that should be taken into account with the

introduction of 5G in Europe? What are the policy and regulatory lessons to be applied for a

successful deployment of 5G in Europe? What does 5G have in common with previous

generations and where is it different? What are the implications? Moreover, is the path towards

the future predetermined by the previous generations, by a prevailing industry structure, or are

there alternatives routes? Is there possibly a fork in the road ahead that requires special


attention from policy makers and regulators, as it may lead to different futures?1 When there

are different futures with different outcomes, is one future more desirable than the other?

To respond to these questions, this research report provides first an assessment of the previous

generations of mobile communication technologies and derives the policy and regulatory

lessons against the backdrop of European leadership in GSM. Secondly, it provides a description

of 5G, the performance objectives that have been set, its architecture and key features and

compares this with previous generations. Thirdly, it describes two possible stylised images for

the future of 5G, an ‘evolution’ image and a ‘revolution’ image. These two images represent two

extremes to capture the widest range of possible 5G futures. These images are deliberately

chosen to represent extremes, as it is not the intention to try to predict the most likely future of

5G. Furthermore, these images of the future do not aim to represent the complexity of how the

actual future may unfold, nor should they be considered as scenarios, such as those initiated by

Shell in the eighties. They are aimed at stimulating the debate on the best set of policy and

regulatory conditions for the successful development and deployment of 5G in Europe.

The ‘evolution’ image follows the pattern of previous generations and current trends. The

‘revolution’ image represents a clear break with the trends as it exploits the opportunities of

open access APIs being enabled by network virtualisation as an architectural foundation of 5G.

These open APIs allow the market entry of a multitude of virtual mobile network operators.

VMNOs dedicated to serve particular industry verticals or economic sectors with tailored feature

sets and tailored qualities of services. These VMNOs may originate from the industries they

serve, such as internal ICT departments extending their reach to customers, from services firms

specialised in and dedicated to a particular industry, from incumbent2 operators diversifying

beyond the mass market of consumers and from start-ups.

These two stylised images reflect two different futures of 5G, two extremes. They yield two

different sets of outcomes that are enabled by two different sets of policies and regulatory

interventions. They constitute a fork in the road that policy makers and regulators will have to

navigate in 2017.

This research report is structured as follows: in Section 2 the European leadership role in mobile

communications is explored. It also derives the regularities across the subsequent generations

1G through 4G and provides an interpretation in the light of the next generation, i.e. 5G. Section

3 describes the architecture and features of 5G, with special attention to virtualisation. The

demand side expectations are also captured in this section. Section 4 introduces the two stylised

images of the future of 5G. Section 5 describes the ‘Evolution’ image using the Porter/Wheelen

industry structure dimensions and includes a sketch of the anticipated industry outcome. This

outcome is compared with the GSM success factors identified in Section 2. In Section 6, the

1 The metaphorical ‘fork in the road’ does not suggest there are only two futures.

2 The term ‘incumbent’ is used to denote mobile network and service operators as they exist at the time or in the time

period as referenced. The term does not typically include mobile virtual network operators.


policy and regulatory actions are derived that would enable the ‘Evolution’ image. Section 7

describes the ‘Revolution’ image and the anticipated outcome, while Section 8 captures the

policy and regulatory actions that would be required to enable the image. Section 9 provides as

a summary an overview of the pros and cons of the two stylised images. As background

information, Annex A provides a short brief on the characteristics of the mobile communications

business. Annex B provides the list of abbreviations and acronyms. Annex C presents a timeline

of major developments in mobile communications. Annex D explores the 5G related radio

frequency management challenges and Annex E addresses net neutrality in the context of

managed services.


2 The leadership role

In describing the European leadership in mobile communications typically, reference is made to

the global success of GSM, a second generation technology introduced in 1991, which reached

its peak in deployment in 2015 with 3.83 billion subscribers through over 700 operators in 219

countries and territories. This is phenomenal achievement, especially when recognising that the

nearest competing technology – CDMA – reached its peak at 374 million subscribers also in

2015. This represents a factor 10 difference.

However, the next generation 3G – UMTS – is generally being considered as less successful,

having had a slow start in deployment compared to a much more rapid uptake in the USA and

Asia. Nonetheless, from a consumer welfare perspective, 3G can be considered as quite

successful, considering the price levels and the additional functionality provided.

For an appreciation of the differences a comparison is made between 2G and 3G based on the

‘roadmap to market’ and the ‘leadership role’ as identified for GSM by Hillebrand3 (see Table 1

and Table 2 below).

Table 1: The road map to market: 2G and 3G compared

Legend: Action similar to 2G

Different action compared to 2G, but conducive

No similar action

2G – GSM 3G – UMTS

The top plane – political level to generate the

political will to make an agreement on GSM

happen:

Agreement between the French and German

Heads of State of November 1984 and the

commitment of the UK in 1986

No similar political level engagement by Member

States

Opening up of a new range of frequencies Similar action with 24% more bandwidth being

allocated

Linking the release of new spectrum to the

market with a new technology

Similar action through auctions; freeing up

previous allocations by introducing technology

neutral assignments

EC Directive to reserve the frequency bands for

the GSM technology

The EC Directive on a timely assignment process;

no threat of alternative standards being

considered for deployment; large installed base of

3 Source: Hillebrand (2002).



2G/2.5G

The second plane – obtaining the commitment of

the cellular radio operators to purchase the new

networks and open a service on a common date

At least three large markets had to come on

stream in the same time to generate the desired

economies of scale

Timely assignment of licenses including major

markets; deployment delays due to economic

setback in the aftermath of the telecom/internet

bubble

Competitive pressure was required to drive

volume

Highly competitive market; being depressed in

the aftermath of the telecom/internet bubble

Use of a common standard allowing for new

revenues from international roaming at almost

zero incremental costs

No change in market structure, no new gains in

moving to the next generation

The third plane – the technical standardisation

effort

Focusing the R&D efforts of the supply industry Preceded by EU R&D program, standardisation

process in ETSI; participants changed from only

European to becoming global in 3GPP

Providing mediation between buyers and

suppliers of networks

Similar situation; buyers and suppliers changed

from predominant European to become global

The fourth plane – the industrialisation by the

supply industry

To be able to recognise the market and its size to

have the confidence for the deep investments

required

Expectation regarding the mobile internet are

very high during the euphoric period and turn

negative after the bubble burst, just after the first

major licenses have been awarded

Semi-conductor industry to be pulled behind the

equipment manufacturers

The semi-conductor industry is aligned, but

impacted by the telecom industry set back

Source: Authors


Table 2: Inputs to the leadership role: 2G and 3G compared

Legend: Action similar to 2G

Different action compared to 2G, but conducive

No similar action


Technology development efforts of France,

Germany, Sweden and Finland

National government-led R&D is replaced by EU-

coordinated and co-funded R&D; this dilutes the

relationship to national industrial interests and

policies, but fits the EU model

Efforts of the French and German operators to

plan a next generation system for a mass market

No new addressable market is created, no similar

transition applies; but installed base could be

leveraged

Very positive market take-up of cellular radio

services in the Nordic countries

The prospect of mobile internet drove demand

expectations strongly

Effort that had to be made by the DTI to bridge

between its European partners and its domestic

competitive players Cellnet and Vodafone

Strong competition was typical for all national

markets in Europe

A shrewd move by the Commission to table a

directive on safeguarding the frequency bands for

a Pan-European cellular radio system;

Such a move was not needed in the 3G context

Close working relationship that the GSM group

achieved between key national officials

The European project changed the role of

national officials, shifting it from inter-state to EU

level

A slice of good luck and well-judged timing A slice of bad luck in terms of how the timing

turned out

Source: Authors

As Ungerer observed,4 the deployment of GSM and DCS1800 systems in Europe was unique

because it coincided with the de-monopolisation and introduction of competition in mobile

communications. The early accelerated mass deployment of GSM was mainly due to new

entrants. At the end of 1993, digital was accounting for only 9% of mobile terminals and new

entrant Mannesmann D2 in Germany accounted for 46% of the European GSM market. With the

competitive pressure from Mannesmann on Deutsche Telecom, the German market

represented 79% of the digital market in Europe.

4 Source: Private conversation in the context of this project.


From 1993 to 1996, a number of procedures were undertaken under EU competition law to

force fair terms for new entrants in several countries, including Italy, Spain and Ireland. This

culminated in the Mobile competition directive issued in 1996, which also mandated the issuing

of the DCS1800 licenses with the deadline of 1 January 1998 – coinciding with the date on which

the European telecommunications market was to be liberalised.

In 1998 there was a debate in the Commission on whether the licensing of 3G should again be

mandated under competition law or under the sector-specific internal market regulation.

Competition law would have given a much more direct enforcement role to the Commission in

overviewing and coordinating licensing and auctions. In the end, the decision was taken in

favour of the internal market approach on which the electronic communications regulatory

framework is based.

With 3G auctions construed to maximise proceeds for the national budget and the auctions

coinciding with the Internet bubble, meant that investment resources were mainly spent on

licenses. New entrants were no longer the main drivers and deployment was mainly with

incumbents and GSM entrants of the nineties. All spent large amounts on the 3G licenses to

secure their 2G position, not for the rapid deployment of 3G. All of this led to a slow deployment

and the loss of the European position in digital mobile.

It should also be noted that the context has changed significantly between the launch of 2G and

the launch of 3G, and further into the 4G era: (1) the market has been fully liberalised and has

become highly competitive; (2) the position of European equipment manufacturers and mobile

operators has changed as the industry has become global; (3) the role of nation states has

changed as part of the liberalisation process and as part of the European Union project; (4) 2G

was instrumental in establishing the mass consumer market, while 3G and 4G are largely

representing replacement markets for voice and enhancement markets for data; and (5) the role

of the device market has become much more important, the choice of smartphone and related

applications platform have become leading in the decision making process of consumers.

As Fejióo et al. pointed out, during the earlier mobile generations a ‘virtuous circle’ of

investment, innovation and adoption of services had been in play. With the introduction of 4G,

this cycle appears to be broken, being replaced by a cycle that runs in the opposite direction.

Now, the innovation and adoption of services require investments from mobile operators

although these will not necessarily lead to an increase in operators’ revenues.5

Hence, actions that were identified as having been crucial to the leadership role in GSM have to

be reinterpreted in the current context of 5G.

5 Source: Chapter on Spain by Feijóo, Gómez-Barroso, Coomote and Ramos in “The dynamics of broadband markets in

Europe – Realizing the 2020 Digital Agenda” by Lemstra & Melody (2015).


2.1 Regularities in next generation mobile communication: 1G

through 4G

Notwithstanding the differences between the generations, many actions are part of a recurring

pattern, a pattern typical for the introduction of a new generation of mobile technology. For

Europe, these events and actions – generally called attributes – have been captured in Table 3

for the generations 1G through 4G. The column 5G has been added to capture those attributes

as they could be observed to date.6

It should be noted that GSM not only represented a major growth phenomenon, it also

established the foundational elements in the cellular communications business that are still valid

today in Europe, such as calling party pays, international roaming and mutual recognition of

terminal devices. In Table 3, these have been denoted as ‘established routine’ (est. rout.).

Table 3: Recurring pattern 1G through 5G

Attributes

1G – NMT 2G – GSM 3G – UMTS 4G – LTE 5G

Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date

Initiative for next

generation

development

Televerket +

Nordic

Incumbent

operators

1970 UK+FR

incumbent

operators

1981 Govt.

representat

ives in

FAMOUS

1991 3GPP study

into LTE

LTE-

Advanced

2004

2012

EC FP7

METIS

initiative;

ITU WP 5D

2011

2011

Research into

next generation

requirements

and technology

Televerket +

Nordic

Incumbent

operators

FT + DT,

incumbent

operators

1984 RACE 1

RACE 2

ACTS

1985

1990

1995

METIS

5GPPP

METIS-II

2012

2013

2015

R&D

collaboration

agreements

EU – South

Korea w/

Japan w/

China w/

Brazil

2014

2015

2015

2016

Global set of

requirements for

the next

generation

ITU

IMT-2000

1999 ITU

IMT-

Advanced

2008 ITU IMT for

2020 and

beyond

2012

Global allocation

of mobile bands

ITU ITU-WARC 1979 ITU-WRC 2000 ITU-WRC ITU-WRC

targets 400

MHz; ITU

WRC to

specify

2015

2019

Allocation of

additional

CEPT CEPT 1982 CEPT CEPT CEPT

proposal

2015

6 See also Annex A for a high level description of the cellular communications business and Annex C for a timeline of

major events in the communications industry.


Attributes


Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date

spectrum 24.5-27.5

31.8-33.4

40.5-43.5

GHz;

decision

2017

Regional

harmonisation of

spectrum for

dedicated

standard

EC Directive 1987 Est. rout. Est. rout. EC 5G

Action Plan

2017

Newly allocated

band(s) (MHz)

450 GSM 900

GMS-R 890

GSM450

2100 800 2100

2600 3400-

3800 700

24.5-27.5

31.8-33.4

40.5-43.5

60 GHz

Amount of

spectrum

allocated (MHz)

GSM 2x25

DCS 1x75

GSM-R 2x 4

1982

1993

2006

155 MHz 60 120 190

400 60

Tbd in

WRC2019

Assignment

method

Assignment Assignment;

Beauty

contest

Auction;

Beauty

contest

Auction

Political

endorsement

Quadripartite

agreement

EC Directive

on use of 900

MHz

1986

1987

3G Green

Paper, intro

2000

Endorse-

ment UMTS

Forum

1993

1995

EC

Directive

on 700

MHz

SDO and start

standardisation

NMT:

Televerket

and Nordic

operators

1975 CEPT (1989)

3GPP (>1999)

EC-GSM-IoT

Dec

1982

2015

ETSI

3GPP

1996

1999

3GPP

MTC

2013

3GPP RAN 2015

Participants in

SDO WGs

CEPT:

Operators

CEPT:

Operators;

ETSI:

Operators,

manufact-s,

academic

inst.

Operators,

manufactur

ers,

academic

inst.

Operators,

manufact-s,

academic

inst.

Operators,

manufact-s,

academic

inst.

Country of origin

participants in

SDO WGs7

Europe GSM900:

Europe;

GSM1900:

Europe +

USA Europe

+ Japan

Global Global

7 In this dimension, it is important to recognise the deployment of GSM in countries outside Europe and hence the

inclusion of actors from these countries in the standardisation efforts.


Attributes


Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date

USA Global

Determination

basic parameters

CEPT GSM#13 1987 UMTS Task

Force input

to ETSI,

final by

ETSI

1996

1997

1998

Selection of radio

interface

1987 1998 Above 24

GHz

Expect.

2018

Decision on the

core network

Not

applicable

Replace 1G

circuit

switched core

Retain 2G

circuit and

packet

switched

core

Replace 3G

core by

packet

switched

core; slicing

New radio

interface;

core to be

replaced;

virtualisa-

tion

First release

specification

For

tendering; for

roll-out

1988

1990

First

release

R99;

For service

offering

1999

2000

Mar

3GPP

Release 8

2008 EC Action

Plan target

3GPP R14

target R15

target R16

2019

2017

2019

2020

Entity for

commercial &

operational

coordination

MoU

Association

1987 UMTS

Forum

GSMA

1996 GSMA GSMA

MGMN

Coordination of

introduction;

target date

Operators

through GSM

MoU; 1991

1987 EC

Directive

on licensing

process

with

execution <

Jan 2000

1999 EC 5G

Action Plan

Early intro

Large scale

2018

2020

First spectrum

assignment(s)

Finland

1999

Mar

France 700

MHz

Germany

700

Last spectrum

assignment(s)

Denmark

2001

Sept


Attributes


Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date Key actor/

entity

Date

First commercial

service

NMT450

TACS

1981

1985

GSM900

DCS1800

PCS19008

GSM8005

1992

1993

Sept

1995

Nov

2002

WCDMA

UMTS

2001

2002

Norway

and

Sweden

2009

Availability

terminals

Nokia

Ericsson

Handhelds 1992 PCMCIA

Handsets

2001

2002

Mutual

recognition

Through type

approvals

RTTE

Directive

1999 est. rout. Est. rout.

First roaming

agreement

Telecom

Finland +

Vodafone-UK

June

1992

est. rout. Est. rout. Est. rout.

First million(s)

users

1 mln

10 mln

100 mln

1 mln

10 mln

100 mln

1993

1995

1998

1 mln

100 mln

1000 mln

2003

2006

2012

1 mln

100 mln

1000 mln

2010

2013

2015

1 mln

100 mln

1000 mln

First non-EU

operator

Australia 1993 est. rout. Est. rout. Est. rout.

First major

upgrade

specification/

services (x.5G)

Packet data

(GPRS)

Enhancement

1998

2000

IMS High-

speed

packet

access

HSDPA

HSUPA

2001

2005

2007

Release 10

LTE-

Advanced

2011

Peak deployment NMT 1996 GSM

CDMA9

2015

2015

First retirement Telia

Finland

Dec

2002

Macau June

2015

Last retirement 2010

?

2030

?

Source: Authors.

8 Related to deployments outside Europe.

9 CDMA added for comparison purposes.


2.2 Interpretation of the regularities and trends leading to 5G

The four subsequent generations of mobile network technology show a clear pattern in terms of

succession: every 10 years, a new generation is introduced.10 The developments to date with

respect to 5G are at large aligned with these regularities. Hence, we may expect 5G to be

introduced around 2020.

The initiative for a next generation typically emerges at the time a previous generation is being

introduced in the field, i.e. some 11-10 years before the launch date. In 1G and 2G these

initiatives originated with the mobile operators, at that time the government-owned national

telecom monopolies. These initiative included R&D into the next generation by the incumbent

players. With the introduction of competition, starting with the deployment of 2G and being

fully established when 3G was introduced, the emphasis had shifted to pre-competitive R&D

programs initiated and sponsored by the European Union with participation of equipment

manufacturers, operators and academic research centres. During the 3G era, the operators

typically reoriented their R&D activities towards service provision, while leaving network-related

R&D to the equipment vendors. The 5G-oriented research within the EU funded FP7 and Horizon

2020 programs aligns with this trend, in terms of timing, content and industry participation. The

strategic collaboration agreements on 5G R&D made by the EU with Japan, Korea, China and

Brazil are consistent with 5G to become a global standard, with 3GPP as the standardisation

platform. A platform that was established based on European initiative recognising the extended

geographical scope of the standardisation efforts, largely as a result of the global deployment of

GSM.

At the time of a next generation initiative, the allocation of new frequencies is also made by the

CEPT, in line with agreements made at the global level within ITU-R.11 For the first three

generations, new radio frequency bands were typically found at higher frequencies, which

provided for higher data rates. This nicely coincided with the need for increasing data rates per

user. With increasing mobile use, the pressure for more spectrum mounted and through the

transition from analogue to digital broadcasting, lower frequency bands were becoming

available, e.g. the 800 MHz band as part of 4G. The plans for 5G are in line with this trend, i.e.,

high-end extensions are foreseen in bands between 24 GHz and 83 GHz,12 as well as a low-end

re-allocation of the 700 MHz band.13

However, the linkage between next generation and new spectrum assignments appears to have

become weaker. On the one hand, auctions are organised at the national level as and when new

10 Note that 4G – LTE was ahead of ‘schedule’ with close to 2 years, apparently to stay ahead of WiMAX, which had

become an IMT2000 family member. 11

At certain instances the ITU-R has been leading, at other times the CEPT proposed an allocation scheme to the

WRC. 12

See for details https://www.itu.int/dms_pub/itu-r/oth/0c/0a/R0C0A00000C0014PDFE.pdf 13

In a number of countries the 700 MHz band is made available earlier for use by LTE. See also Annex D.


or re-allocated spectrum becomes available and, on the other hand, spectrum is now assigned

on a technology-neutral basis, i.e. next generation equipment may be deployed in bands

originally assigned for previous generations.14 This trend started with 4G and also applies to 5G.

The prominent coordinating role of operators in the introduction of 2G, in terms of timing and

functionality, has moved to the background in 3G and 4G. On the one hand, the competitive

market is expected to drive the introduction process – and coordination could be interpreted as

collusion – and, on the other hand, the operational aspects of a next generation are now

addressed by the GSMA, the institutional successor of the Memorandum of Understanding

between operators in the 2G era.

At the time of 1G and 2G, the introduction of a new generation required deep investments in

the roll-out of new infrastructure replacing the previous generation. With the deployment of a

packet-overlay network in the form of GPRS, an inter-generational upgrade was introduced:

2.5G. With 3G, the investment in a new radio access and new core equipment became

separated in time: first, a new radio was introduced (Wide band CDMA), which was made

interoperable with the existing 2G circuit switching and 2.5G packet switching core. As part of

3.5G the packet capabilities were upgraded towards HSPA. In 4G, the circuit core was

abandoned and the packet core was further enhanced. Still in 4G, a new modulation technique

was applied on the radio access (OFDMA), requiring upgrades of base stations and handsets,

while earlier generations remained backward compatible with existing evolved core network.

LTE-Advanced, which is providing higher data rates based on carrier aggregation, represents the

inter-generational upgrade to 4.5G. The envisioned evolution towards 5G includes adding a new

radio access in the frequency bands above 24 GHz to be compatible with the existing evolved

packet core (EPC). The plans also project the introduction of virtualisation (Software Defined

Networks and Network Function Virtualisation), which means a further move of functionality

into software and the application of bulk-standard Ethernet switches and computing resources.

This is expected to be a gradual process, starting with new interfaces being added to existing

network equipment.

Evolution of handsets

The replacement model of 1G by 2G implied the need for new devices. With the allocation of

additional GSM bands (1800 and 1900 MHz), followed the introduction of multiband radios

allowing for interoperability within a single generation across multiple frequency bands. With

more bands being allocated and assigned over time, the support of multiple bands by handset

providers in line with the national band plans has become a critical issue. Handset roll-out plans

are being optimised based on device market size and market priorities as perceived by handset

14 While spectrum bands may have been made technology neutral some aspects, such as the channel width, may have

to be aligned with a particular generation of technology. This may involve adaptation of regulatory conditions.


vendors.15 Furthermore, handset functionality extends well beyond mobile network

functionality and the launch of new devices is increasingly driven by smartphone vendors, with

typically a new release every 1-2 years.16 In certain markets, the 3.5 GHz band has been assigned

but has remained unused, lacking appropriate terminal devices.17 Also, the use of carrier

aggregation as part of 4G is subject to terminal-network compatibility. In the evolution towards

5G, this is expected to remain an issue of concern, suggesting the necessity for further

coordination of frequency plans by national administrations.

Evolution of the core network based on standards

Over the generations, the scope of mobile communication standards has evolved from being

national, through being regional to becoming global. That process has been strongly influenced

by the regional and subsequently global success of GSM. With GSM deployed in all regions, it

brought together the interests of operators across the globe in relation to the next generation

standard to be deployed. With 3G being designed to be compatible with the previous

generations, three regional standards resulted. 4G in casu LTE and LTE-Advanced have been

conceived as global standards and are now accepted and deployed as such. 5G will become the

next global standard for mobile communications. Based on its experience with 3G and 4G, the

3GPP as standard development organisation is set to create the 5G specifications.

Evolution towards verticals

While oriented towards the mass market of consumers, GSM has evolved to support a first

public sector vertical market: GSM-R serving European railway operators, for which a separate

frequency band had been allocated in Europe. The GSM-R functionality has become part of the

general GSM specification, such that the functionality was available to address other similar

niche markets. A second public sector vertical is being accommodated as part of 4G release 13

through 15: the public protection and disaster relief (PPDR) sector, which includes the police,

fire brigade and ambulance services. In Europe, the PPDR sector was previously served through

a dedicated system called TETRA, operating in a dedicated band. The sector has concluded that

for the transition from narrowband to broadband it will have to rely on LTE and LTE-Advanced,

as a dedicated broadband system is not a viable option.

15 This has for instance led to the Apple iPhone 5 at release not being compatible with the assigned 4G frequencies in

Belgium. Source: Van der Wee, Verbrugge & Laroy (2015). 16

See Annex C for the introduction dates of Apple iPhone releases as example. 17

See for instance gsacom.com on the limited availability of LTE devices in band 42/43.


3 5G architecture and features18

5G represents a next step in the technological evolution of mobile communications networks:

1G was dedicated to telephony. 2G started as capacity expansion for telephony, to which a

packet-switched overlay network (GPRS) was later added to provide access to the Internet. 3G

was designed for voice and high-speed data communication (implemented through resp. circuit

switching and packet switching). High demand for Internet access accelerated the transition to

the next generation of mobile technology – 4G – also known as Long-Term Evolution (LTE),

which is packet-switched only.19 The upgrade to LTE Advanced, introducing data rate

enhancement through carrier aggregation, was first introduced in 2013. It provides a peak cell

capacity of 1.2 Gbit/s.

3.1 5G requirements

In 2012, the METIS research project, one of many projects dedicated to the development of 5G

within the EU co-funded FP7 and Horizon 2020 research programs, set out the design targets for

5G as follows:

1000 times higher overall capacity 10-100 times more devices

10 to 100 times higher end-user data rates 5 times lower latency

10 times longer battery life

The 1000-fold capacity increase is foreseen to be achieved through 3 simultaneous approaches:

network densification, providing 50x improvement; the use of more spectrum, including higher

frequencies, such as mm Wave (e.g. 24 and 60-80 GHz), providing 10x improvement; and

realising an increase in spectral efficiency, providing 2x improvement.20

In addition to the EU research initiatives, a 5G public-private partnership, called the 5G-PPP, has

been formed. It brings together research institutes, operators and vendors, and was endorsed

by the European Commission. A 5G Infrastructure Association was also founded and has

formulated a vision on 5G including (much similar) high-level requirements (5G Infrastructure

Association, 2015).21.

18 This Section draws on the research first reported in “Imagine 2025” published as Appendix 2 to the CERRE report

“An integrated regulatory framework for digital networks and services” (De Streel & Larouche, 2016). 19

In the context of LTE telephony, services are provided through Voice-over-LTE (VoLTE) or a fall-back to 3G or 2G, so-

called Circuit Switched Fall-back (CSFB) until VoLTE is made available. 20

This compares well with a doubling of aggregate network capacity every 3 years over the last 30 years (Rysavy

Research, 2015). 21

For an overview of global 5G initiatives, see the report by 4G Americas (2014a).


According to the Association 5G Vision’s statement, the 5G design is aimed at:

• bringing together the various radio access technologies (e.g. GSM, UMTS, LTE, Wi-Fi and satellite) to provide the end-users with seamless handovers;

• to provide a multitenant environment for various users groups (mobile operators, broadcasters, public safety and disaster relief, providers of cellular service for the

railways); thereby

• paving the way for virtual pan-European operators, relying on national infrastructures.

The performance objectives formulated are:

• radically higher wireless area capacity (1000x relative to 2010); • much lower round-trip delays (latency


Figure 1: 5G use case families and related examples

Source: NGMN Alliance (2015).

Next to functional requirements related to data rates, latency, number of devices, etc. a set of

design principles have been formulated by the Next Generation Mobile Network Alliance on

behalf of its members. These design principles reflect the operational requirements of the

mobile network operators (see Figure 2).

Figure 2 : 5G design principles



3.2 5G architecture and virtualisation

The overall 5G architecture as foreseen by the NGMN is reflected in Figure 3. It reflects the

layered structure including virtualisation and the use of APIs.

Figure 3: 5G architecture


The major new technological development affecting 5G is network virtualisation and the use of

application programming interfaces (APIs). Network virtualisation refers to implementing the

functions of the communications infrastructure in software running on commercial ‘off-the-

shelf’ computing equipment, essentially Ethernet switches linked by optical fibers being

centrally controlled by software. This follows the virtualisation of data centres and the use of a

modified version of the Internet protocol adapted towards centralised network control. More

specifically, 5G will be implemented based on software-defined networking (SDN) and network

function virtualisation (NFV), mobile edge computing (MEC) and fog computing (FC), in essence

an architecture based on “cloud” computing, linking together a diverse set of resources for

transport, routing, storage and processing, including (user) resources at the edge of the

network. Moreover, it supports the development of new services through application

programming interfaces.22

22 Sources: Patel et al. (2014); 5G Infrastructure Association (2015). For small scale application and experimentation

with virtual networks see for instance the PhD by Strijkers (2014). For information on SDN in general, see Göransson

& Black (2014) and Stallings (2016).


Virtualisation already started in the fixed network with AT&T being in the lead and Verizon a

close follower. AT&T described the motivation to move towards network function virtualisation

(NFV) as follows: “AT&T’s network is comprised of a large and increasing variety of proprietary

hardware appliances. To launch a new network service often requires adding yet another

variety, and finding the space and power to accommodate these boxes is becoming increasingly

difficult. This difficulty is compounded by increasing costs of energy, capital investment, and

rarity of skills necessary to design, integrate and operate increasingly complex hardware-based

appliances. Moreover, hardware-based appliances rapidly reach end-of-life, requiring much of

the procure-design-integrate-deploy cycle to be repeated with little or no revenue benefit.

Additionally, hardware lifecycles are becoming shorter as technology and service innovation

accelerates, and this can inhibit the expeditious roll out of new revenue earning network

services and constrain innovation in an increasingly network-centric connected world. NFV aims

to address these problems by evolving standard IT 29 virtualisation technology to consolidate

many network equipment types onto industry standard high volume servers, switches and

storage that can be located in data centres, network PoPs or on customer premises. This

involves the implementation of network functions in software, called Virtual Network Functions

(VNFs), that can run on a range of general purpose hardware, and that can be moved to, or

instantiated in, various locations in the network as required, without the need for installation of

new equipment.”23 AT&T senior management announced as target 75% of the network to be

virtualised by 2020.

The compelling reasons for applying virtualisation are: lower capital expenditures, benefiting

from economies of scale in the IT industry; lower operating costs; faster deployment of new

services; energy savings; and improved network efficiency.

The European Telecommunications Standards Institute (ETSI) has standardised the framework,

including interfaces and reference architectures for virtualisation (see Figure 4 showing the ETSI

framework, in which virtualised network functions – VNFs – are the nodes or applications by

which operators build services). Other standards and industry groups involved include 3GPP, The

Open Network Foundation, OpenStack, Open Daylight, and OPNFV.24

23 AT&T (2013).

24 4G Americas (2014b).


Figure 4: ETSI ISG network virtualisation framework

Source: Rysavy Research (2015).

The core network, consisting of fewer nodes, provides an easier starting point for virtualisation.

Although more complex, virtualisation of the RAN is expected to provide the greatest network

efficiency gains, particularly for small-cell deployments.25

Virtualisation and the decoupling between radio access technologies (RATs) and the core

network (CN) functionalities support the principle of network slicing. In that way, the various 5G

use cases with different requirements on the radio interface and in terms of data processing in

the core network can be combined and supported by one integrated mobile network. For an

illustration see the ‘paring’ of RATs with CN slices in Figure 5.

25 For an insightful description of virtualisation in particular network slicing, see the 5G Americas White Paper

“Network Slicing for 5G networks and services.” Source:

www.5gamericas.org/files/3214/7975/0104/5G_Americas_Network_Slicing_11.21_Final.pdf Retrieved: 2016-11-21.


Figure 5: Network slice examples

Source: 5G Americas (2016).

3.3 5G roadmap

The high level roadmap for the various 5G related activities is reflected in Figure 6. Note that the

functionality foreseen for 5G will become available over time in a series of releases of the

specifications.


Figure 5: 5G roadmap, 2014-2024

Source: 5G Infrastructure Association (2015).

3.4 The demand-side perspective

This section provides the demand-side perspective of 5G, largely as an extension of the current

trends.

3.4.1 The market for connections and devices

Figure 6 provides Cisco’s forecast for growth and penetration of global consumer mobile

services towards 2019 (Cisco, 2015b).


Figure 6: Forecast global mobile consumer services, 2019

Source: Cisco VNI Mobile, 2015(Cisco, 2007).

Note that all but one of the services shown are applications which use the mobile infrastructure

to obtain access to the Internet. Only MMS is an integrated service, which is shown with a

negative growth rate. Moreover, mobile telephony and SMS as distinct services have

disappeared from the radar screen, having become part of mobile social networking.

Using Cisco’s VNI 2016 projections, the mobile communications landscape will have the

following features by the time 5G is introduced, i.e. 2020:

• Global mobile data traffic will increase nearly eightfold between 2015 and 2020. Mobile data traffic will grow at a compound annual growth rate (CAGR) of 53 percent from 2015

to 2020, reaching 30.6 exabytes26 per month by 2020.

• By 2020, there will be 1.5 mobile devices per capita. There will be 11.6 billion mobile-connected devices by 2020, including M2M modules—exceeding the world’s projected

population at that time (7.8 billion).

• Mobile network connection data rates will increase more than threefold by 2020. The average mobile network connection speed (2.0 Mbit/s in 2015) will reach nearly 6.5

Mbit/s by 2020.

• By 2020, 4G will represent 40.5 percent of connections and 72 percent of total traffic. By 2020, a 4G connection will generate 3.3 times more traffic on average than a non-4G

connection.

• By 2020, more than 60 percent of all devices connected to the mobile network will be “smart” devices. The vast majority of mobile data traffic (98 percent) will originate from

these smart devices by 2020, up from 89 percent in 2015.

• By 2020, 66 percent of all global mobile devices will be capable of connecting to an IPv6 mobile network. There will be 7.6 billion Ipv6-capable devices by 2020.

26 Exabytes: 1 EB = 1000

6 bytes = 10

18 bytes = 1000 petabytes = 1million terabytes = 1billion gigabytes.


• By 2020, 75 percent of the world’s mobile data traffic will be video. Mobile video will increase 11-fold between 2015 and 2020.

• The amount of mobile data traffic generated by tablets by 2020 (2.6 exabytes per month) will be 7.6 times higher than in 2015, a CAGR of 50 percent.

• The average smartphone will generate 4.4 GB of traffic per month by 2020, nearly a fivefold increase over the 2015 average of 929 MB per month. By 2020, aggregate

smartphone traffic will be 8.8 times greater than it is today, with a CAGR of 54 percent.

• Currently, more than half of all traffic from mobile-connected devices (almost 3.9 exabytes) is offloaded to the fixed network by means of Wi-Fi devices and femtocells

each month. Without Wi-Fi and femtocell offload, total mobile data traffic would grow

at a CAGR of 55 percent between 2015 and 2020, instead of the projected CAGR of 53

percent.

• The Middle East and Africa will have the strongest mobile data traffic growth of any region with a 71-percent CAGR. This region will be followed by Asia Pacific at 54 percent

and Central and Eastern Europe at 52 percent.

See also Figure 8 for the projected mobile traffic growth.

Figure 6: Forecast of global mobile traffic growth, 2015-2020

Source: Cisco VNI Mobile, 2016.

In this projection, Western Europe accounts for 9% of the total volume in 2020 and Central and

Eastern Europe for 14%.

Figure 7 shows the projection for connections and devices towards 2020.


Figure 7: Global growth of mobile devices and connections, 2015-2020


Figure 8 reflects the distribution of devices/connections by technology 2G-4G and LPWA. The

percentages refer to the device/connection share of the total.27

Figure 8: Global mobile devices and connections by technology, 2015-2020


3.4.2 The IoT market

Following the major transition from car-borne phones to handsets, the next major expansion of

the addressable market is the Internet-of-Things (IoT), or the interconnection of uniquely

identifiable embedded computing-like devices using the Internet. IoT requires: (1) the transition

27 LPWA: Low power wireless access, mainly used for connections in the context of IoT.


to IPv6, which has a much larger address space of up to 3.4×1038; (2) high as well as very low

data rates; and (3) very low energy consumption.

IoT includes the earlier form of machine-to-machine (M2M) communication, which originated in

the field of industrial instrumentation. The ubiquitous use of the Internet facilitates M2M

communication and expands its range of applications. Previously, this was also denoted as

telematics. Meanwhile, many mobile operators have created business departments dedicated to

providing M2M services. As an example, a number of energy utility companies have outsourced

the collection of smart-meter data to communication providers. At least one of the utility

companies has acquired a radio spectrum license to set up a network to collect metering data

over the air.28

The lowest-cost devices enabling M2M communications today are GPRS modems, which may

become obsolete as operators decommission their GSM systems. HSPA is also used for M2M

communications. Furthermore, LTE has been optimised to efficiently communicate small bursts

of information, making it well suited for M2M. Low-cost LTE modem options included in 3GPP

releases 10 through 13, reduced costs, improved the communications range, and extended

battery life (Rysavy Research, 2015). Figure 9 reflects the forecasted use of mobile technologies

for M2M.

Figure 9: Global mobile M2M connections by technology, 2015-2020


5G is set to serve the two different segments of the IoT market: (1) the market of massive

machine-type communications (mMTC), related to smart cities, smart infrastructures and

objects (sensors and actuators); and (2) the market for ultra-reliable and low-latency machine-

type communication (uMTC), related to autonomous vehicle control, smart electricity grids and

factory cell automation.29

28 Alliander in the Netherlands acquired a license in the 450 MHz band to deploy CDMA450.

29 Osseiran, Monserrat & Marsch (2016).


IoT is considered to include a very wide range of applications such as: environmental

monitoring; energy management; remote health monitoring and notification; building and home

automation; smart vehicles; and more. Contributing to the growing adoption of Internet-of-

Everything (IoE) are wearable devices. Wearable devices have the capability to connect and

communicate to the network either directly through embedded cellular connectivity or through

another device (primarily a smartphone) using Wi-Fi, Bluetooth, or another technology. These

devices come in various shapes and forms, ranging from smart watches, smart glasses, heads-up

displays (HUDs), health and fitness trackers, health monitors, wearable scanners and navigation

devices, smart clothing, etc. The growth in these devices has been fuelled by enhancements in

technology making the devices light enough to be worn. These advances are being combined

with fashion to match personal styles, especially in the consumer electronics segment, along

with network improvements and the growth of applications, such as location-based services and

augmented reality.30

According to Cisco´s 2015 VNI projection, M2M connections will grow to over 10 billion

worldwide by 2019, with 4.6 Petabytes of traffic per month. See Figure 10 for the growth rates

and a breakdown by industry vertical (Cisco, 2015a). In the 2016 outlook, the forecast is lowered

significantly based on a lower take up in the early years: globally, M2M connections will grow

from 604 million in 2015 to 3.1 billion by 2020, a 38-percent CAGR.

Figure 10: Forecast global M2M connections by industry vertical, 2014-2019

Source: Cisco, VNI Mobile 2015.

30 Cisco (2016).


4 The two stylised images of the 5G future

It has been said that ‘forecasting the future is best done in hindsight’. Nonetheless, exploring

what the future might bring remains of critical importance in successfully managing a business.

As the policy enterprise has in common with the business enterprise the need to explore the

future to devise successful policies, we present two contrasting stylised images for the 2020-

2025 horizon. These two images represent two extremes to capture the widest range of possible

5G futures. These images of the future neither aim at representing the complexity of how the

actual future may unfold, nor should be considered as scenarios, such as those initiated by Shell

in the eighties.

To avoid any doubt on the purpose of the images, we are not suggesting that either one of these

represents the most likely future outcome. The future may evolve as a mixture of these two in a

pattern which varies over time and place, or may be different from what is described. The two

images have been developed to highlight the range of 5G challenges which are likely to be faced,

and thus focus attention on the key short and medium term choices concerning policy and

regulation which have to be made to assure the successful development and deployment of 5G

in Europe.

4.1 ‘Evolution’ and ‘Revolution’

The stylised images are called ‘Evolution’ and ‘Revolution’. They represent respectively a

continuation of the development path of mobile communications as it can be derived from the

development of the previous generations, i.e. 1G through 4G, and a break with past

developments made possible through technological developments, i.e. the virtualisation of

communications networking. The contrast between the images is in the two different industry

structures they represent. On the one hand, the continuation of an oligopolistic market

structure of incumbent mobile network and service providers, and on the other a market that is

driven by a wide range of firms specialised in serving the requirements of different (vertical)

industries through applications running on open access network infrastructures, providing

seamless service on a regional basis.

The two images are informed by research into, on the one hand, the development of 1G through

4G – in particular an investigation into regularities and trends that can be observed – and on the

other hand, the relatively recent experience with the development and deployment of

virtualisation in data centres and the steps taken by AT&T and Verizon to virtualise their telecom

infrastructures.


4.2 The industry structure

To describe the 5G communication services industry structure and environment under the two

stylised images, use is made of the framing provided by Porter and Wheelen, i.e. a combination

of the Five Forces framework and the SEPT framework respectively, to which the environmental

dimension is added31 (see Figure 11).

Figure 11: Framework for industry analysis (Porter-Wheelen)

Source: Author, based on Porter (1980) and Wheelen & Hunger (1983).

The SEPT dimensions provide a sketch of the broader socio-economic context in which the more

detailed ‘Evolution’ and ‘Revolution’ images are positioned.

31 It is acknowledged that the Porter framework provides a static view of the industry and needs to be used in a

comparative static mode to capture dynamic aspects. In the context of the image development its main purpose is to

structure the information and act as a check to assure all relevant dimensions are addressed. Sources: Porter (1980)

and Wheelen & Hunger (1983).

INDUSTRYCOMPETITORS

RIVALRY AMONGEXISTING FIRMS

POTENTIALENTRANTS

SUPPLIERS BUYERS

SUBSTITUTES

Bargainingpower ofsuppliers

Bargainingpower ofbuyers

Barrier to entry

Threat of substituteproducts or services

Political

Societal

Techno-

logical

Economic

Threat of entry


5 The ‘Evolution’ image

In the ‘Evolution’ image? the regularities and trends that can be observed from the previous

generations of mobile communication, i.e. 1G through 4G, are considered as the main

determinants of the 5G future. The incumbent operators consider spectrum holdings, the active

parts of the network and customer relationships as their core strategic assets – while passive

infrastructure (e.g. towers) are sold and leased back, and maintenance is increasingly

outsourced to third parties. It provides opportunities for vertical integration of networks and

services and thereby differentiation from the so-called Over-the-Top service providers. The

incumbent operators deploy new technologies to defend and strengthen their position vis-à-vis

the competitors and in developing the relationship with their customers. A key assumption in

this image is that the core business of the mobile operators continues to be serving the mass

market of consumers.

5.1 Anticipated outcome – attractiveness of the outcome32

This section describes the anticipated outcome of the future image in hindsight.

In the ‘Evolution’ image – described in Section 5.2 – the leading players are the incumbent

mobile operators. Given the competitive market place and consumers having become used to

getting access to more bandwidth with each new generation at roughly the same price, the

profit margins remain small. Hence, the incumbents have a strong incentive to optimise past

investments and to be prudent with new investments. The business case has become more

challenging with each new generation, as the investment costs per subscriber increased while

per subscriber revenues remained flat. See Figure 12.33

32 In this Section we compare the outcome of the image of the 5G future with the GSM success factors derived in

Section 2. 33

Source: SMART 2014/0008 Identification and quantification of key socio-economic data to support strategic

planning for the introduction of 5G in Europe. Final report prepared by Tech4i2, Realwireless, Trinity College Dublin

and InterDigital. (2016). Doi: 10.2759/56657.


Figure 12. Estimated per subscriber costs of next generation mobile technologies, 1990 - 2025

Source: SMART2014/0008 (2016)

LTE being a high-capacity All-IP system removed past infrastructure bottlenecks and provided a

controlled path towards the future, with the introduction of LTE-Advanced, as well as upgrades

of functionality through annual releases. With the relative low and stable prices paid by end-

users, largely irrespective of the increase in data rates offered, this evolutionary image fits the

desire for a stable business model with relatively flat investment levels.34 This provides for

relatively stable and predictable performance.

As the 5G architecture evolved by adding new radio interfaces in bands above 24 GHz to the

existing LTE core network, incumbents can serve the demand for higher data rates in an

incremental way, particular in high density city areas, as and when demand is manifest. The

replacement of the LTE core network by a 5G core network could be phased, based on new

products becoming stable and being provided at lower costs.

As the newly available frequency band below 1 GHz, i.e. the 700 MHz band, was already

auctioned for use by LTE and LTE-Advanced, there was no direct linkage between the release of

this new spectrum and the introduction of the new 5G technology. The introduction of spectrum

bands above 24 GHz was, and still is, of importance for network densification to provide higher

data rates. However, this did not provide a window of opportunity for infrastructure market

entry as part of 5G. Hence, increased competition as a major driver of success related to GSM

was lacking in the context of 5G.

The market of IoT provided opportunities for growth, but this market is highly diverse and

served by competing technologies operating in unlicensed bands, such as LoRa. These

34 The investment profile typically reflects investments in coverage in the early period and investments in capacity

upgrades (densification) in the subsequent period.


alternative technologies were designed for IoT from the outset, with long range and low power

as design objectives. These systems appeared to be more effective than scaled-down versions of

high-capacity cellular systems. Hence, they provide effec

Documents

Towards the successful deployment of 5G in Europe · 2020. 12. 14. · Centre on Regulation in Europe (CERRE) asbl Avenue Louise, 475 (Box 10) - B-1050 Brussels - Belgium Ph: +32